Method and System for Optimized Streaming Game Server

1. A system for synthesizing a compressed video bit stream from an interactive three-dimensional (3-D) graphical application, the system comprising: means for processing an interactive application, frame by frame, wherein each frame comprises a plurality of objects; means for interception of a set of graphical rendering commands generated by the processing of the interactive application; means for extraction of a set of geomtric and texture information from the intercepted set of commands per object; means for generating a modified set of graphical commands; means for rendering image frames from the modified set; means for pre-processing of the rendered image frames; means for encoding of the pre-processed image frames; means for controlling the means for generating, the means for rendering, the means for pre-processing, and the means for encoding based on a measure of encoding quality; and means for defining a level of detail of rendered images based on a pre-analysis of the extracted geometric and texture information, wherein the means for defining a level of detail of rendered images comprises a level-of-interest based control that is based on a region of interest segmentation process in which a relative visual quality value is assigned to objects in a respective frame.

2. A system according to claim 1 , further comprising means for computing two dimensional (2-D) visual information from the extracted geometric and texture information.

3. A system according to claim 2 , wherein the 2-D visual information is computed 2-D motion field relative to a reference frame, a level-of-interest classification for each object in rendered image frames and a foreground/background segmentation.

4. A system according to claim 2 , further comprising means for controlling the modified commands based on anyone of computed two-dimensional (2-D) visual information.

5. The system according to claim 4 , wherein the means for controlling the modified commands comprise means for inserting a dedicated level of detail for each object in the rendered image frames.

6. A system according to claim 2 , further comprising means for controlling the pre-processing based on the computed two-dimensional (2-D) visual information.

7. A system according to claim 2 , further comprising means for controlling the encoding modules based on the computed two-dimensional (2-D) visual information.

8. A system according to claim 7 , further comprising means for controlling entropy encoding, quantization and motion estimation, based on the computed two dimensional (2-D) visual information.

9. A system according to claim 1 , further comprising means for controlling the encoding means based on an encoding quality criteria.

10. The system according to claim 1 , further comprising: means for defining a level of detail of rendered images based on a pre-analysis of the extracted geometric and texture information; and means for modifying the defined level of detail of the rendered images.

11. A system for synthesizing a compressed video bit stream from an interactive three-dimensional (3-D) graphical application, the system comprising: an interception device configured to intercept a set of graphical rendering commands generated by processing of the interactive 3-D graphical application; an extraction device configured to extract a set of geometric and texture information from the intercepted set of commands per object; a processing device configured to generate a modified set of graphical commands; a rendering device configured to render image frames from the modified set; a pre-processing device configured to pre-process the rendered image frames; an encoding device configured to encode the pre-processed image frames; a feed-back control device configured to control the processing device, the rendering device, the pre-processing device, and the encoding device based on a measure of encoding quality; and a feed-forward control device configured to define a level of detail of rendered images based on a pre-analysis of the extracted geometric and texture information, wherein the feed-forward control devices comprises a level-of-interest based control that is based on a region of interest segmentation process in which a relative visual quality value is assigned to objects in a respective frame.

12. The system according to claim 11 , wherein the feed-back control device is further configured to modify the defined level of detail of the rendered images.

13. The system according to claim 12 , wherein the feed-back control device is configured to modify the defined level of detail of the rendered images based on a present measure of compression quality.

14. The system according to claim 13 , wherein the feed-back control device is configured to modify the defined level of detail of the rendered images based on an available stream bandwidth.

15. The system according to claim 11 , further comprising a second processing device configured to process an interactive 3-D graphical application, frame by frame, wherein each frame comprises a plurality of objects.

16. A method for synthesizing a compressed video bit stream from an interactive three-dimensional (3-D) graphical application, the method comprising: processing an interactive 3-D graphical application, frame by frame, wherein each frame comprises a plurality of objects; intercepting a set of graphical rendering commands generated by the processing of the interactive 3-D graphical application; extracting a set of geometric and texture information from intercepted set of commands per object; generating a modified set of graphical commands; rendering image frames from the modified set; pre-processing the rendered image frames; encoding the pre-processed image frames; controlling, via a feed-back control device, the generating, the rendering, the pre-processing, and the encoding based on a measure of encoding quality; and defining, via a feed-forward control device, a level of detail of rendered images based on a pre-analysis of the extracted geometric and texture information, wherein the feed-forward control devices comprises a level-of-interest based control that is based on a region of interest segmentation process in which a relative visual quality value is assigned to objects in a respective frame.

17. The method according to claim 16 , further comprising: defining a level of detail of rendered images based on a pre-analysis of the intercepted set of rendering commands; modifying the defined level of detail of the rendered images based on an encoding quality criteria.

18. A method for synthesizing a compressed video bit stream from an interactive three-dimensional (3-D) graphical application, the method comprising: providing an interception mechanism to intercept a set of rendering commands generated by an interactive 3-D graphical application; providing a processing mechanism for the intercepted commands; providing a feed-forward control mechanism configured to define a level of detail of rendered images based on a pre-analysis of the intercepted set of rendering commands, wherein the level of detail is defined according to a region of interest segmentation process in which a relative visual quality value is assigned to objects in a respective frame; providing a feed-back control mechanism from an encoder, wherein the feed-back control mechanism is configured to modify the defined level of detail of the rendered images based on an encoding quality criteria; and modifying the rendered images based on the feed-forward control mechanism and the feed-back control mechanism to optimize compression quality subject to a streaming bandwidth constraint and a system delay constraint.

19. A system for synthesizing a compressed video bit stream from an interactive three-dimensional (3-D) graphical application, comprising: an interception device configured to intercept a set of rendering commands generated by an interactive 3-D graphical application; a processing device for the intercepted set of rendering commands; a feed-forward control device configured to define a level of detail of rendered images based on a pre-analysis of the intercepted set of rendering commands, wherein the level of detail is defined according to a region of interest segmentation process in which a relative visual quality value is assigned to objects in a respective frame; a feed-back control device configured to modify the defined level of detail of the rendered images based on an encoding quality criteria; and programmed instructions configured to modify the rendered images according to data generated by the feed-forward control device and the feed-back control device to produce an optimized compression subject to a streaming bandwidth constraint and a system delay constraint.

20. The method according to claim 18 , wherein the modifying the rendered images comprises at least one of modifying levels of details for different objects within the respective rendered images, modifying post-filtering of the rendered images, or modifying the compression quality of different objects within the respective rendered images.